SSD Health Percentage Is Not the Whole Story: What Used-Laptop Buyers Should Check

A used laptop listing says the SSD has “97% health.” That sounds reassuring, but it is not a complete storage diagnosis. A health percentage can summarize one endurance estimate while hiding other information that matters more immediately: media errors, critical warnings, temperature history, available spare capacity, unsafe shutdowns, controller problems, or an SSD that simply performs badly under sustained load.

NVMe drives expose a standard health log containing several useful fields. Windows can read many of them through the NVMe health interface. The important part for buyers is understanding that “health” is not one universal measurement. Different tools may turn several raw values into a simplified percentage, and that simplified number can be easy to overinterpret.

What NVMe Percentage Used actually means

The NVMe health log includes a field called Percentage Used. Microsoft’s documentation describes it as a vendor-specific estimate of the percentage of the NVM subsystem’s life that has been consumed, based on actual use and the manufacturer’s prediction of endurance.

A value of 100 means the estimated endurance has been consumed. Importantly, Microsoft notes that this does not necessarily mean the SSD has failed. The value can even exceed 100.

That is very different from a simple “100% = dead” interpretation.

Why 97% health and 3% used may be the same underlying idea

Some utilities present remaining life. Others present life consumed. A tool might translate 3% used into 97% health.

That transformation is convenient, but it can make the number look more precise than it really is. The underlying endurance estimate is still based on vendor-specific logic.

Percentage Used is not a failure predictor

An SSD can fail before reaching its rated endurance because of controller faults, firmware issues, power events, manufacturing defects, or other problems. Another SSD can continue working beyond its nominal endurance estimate.

Endurance is therefore one risk dimension, not a countdown timer.

Check the Critical Warning field first

NVMe devices can report critical warning bits for conditions such as low available spare capacity, temperature threshold events, degraded reliability, read-only mode, or volatile-memory backup failure where applicable.

A used drive with no critical warnings is more reassuring than one with active warnings even if both show similar life-used percentages.

Available Spare can matter more than the headline percentage

Flash storage maintains spare blocks that can replace cells that become unusable over time. NVMe reports normalized available spare capacity and a threshold.

If available spare falls below the threshold, the drive can issue a warning. A drive near its spare threshold deserves more attention than a drive with healthy spare capacity even if both still boot and benchmark normally.

Media and data-integrity errors are a major red flag

The NVMe health log includes media and data-integrity error counts.

A non-zero value is not automatically proof that data loss is imminent because context matters. But repeated or increasing media errors are much more concerning than a slightly worn endurance estimate.

For a used-device inspection, record the error count, run a controlled test, and check whether the count increases.

Error-log entries are not the same as media errors

The NVMe health log can also expose an error-information-log entry count. That counter can include controller-recorded events that are not necessarily flash-media failure.

This is one reason raw SMART values need interpretation rather than panic. The pattern matters: critical warning state, media errors, reliability warning, and increasing failure-related counters are more meaningful together than one isolated number.

Unsafe shutdowns tell you about usage history

NVMe devices record unsafe shutdown events. These can occur when power is removed without a clean shutdown.

A high count does not automatically mean the SSD is bad. A desktop can accumulate unsafe shutdowns from power cuts; a laptop can record them after crashes or battery disconnects.

But the value can reveal how the system has been treated. Combined with filesystem problems or other SMART warnings, it becomes more interesting.

Power-on hours add context that health percentage lacks

A drive with 98% life remaining and 20,000 power-on hours has a different history from a drive with the same percentage and 500 hours.

Neither is automatically better. Power-on hours tell you time in service, while endurance counters tell you write-related wear. Both help build a more complete picture.

Data Written can expose heavy workloads

NVMe reports data units written by the host. This can be converted into an approximate total write volume.

A laptop used for office work may have relatively modest writes. A machine used for video editing, virtual machines, scratch disks, databases, or heavy development can accumulate much more.

High write volume is not automatically bad if the SSD is designed for it. It is simply part of the history.

TBW is a warranty-endurance figure, not a death point

SSD manufacturers commonly publish an endurance specification such as TBW—terabytes written—alongside a warranty period.

That number is often tied to warranty terms and expected endurance. Reaching it does not mean the NAND suddenly stops working on the next write. Similarly, being far below TBW does not guarantee the controller or flash cannot fail for some other reason.

Why SSD health tools can disagree

• they interpret vendor-specific attributes differently;
• they show raw NVMe Percentage Used versus calculated remaining life;
• they combine several values into one score;
• they read through different storage drivers or controllers;
• they have incomplete support for a specific SSD model.

If two utilities disagree, return to the raw NVMe or SMART fields and the SSD manufacturer’s documentation.

SATA SMART and NVMe SMART are related but not identical

Older SATA SSDs expose SMART attributes through a different model, often with vendor-specific attribute IDs. NVMe standardized a dedicated health log with fields such as Percentage Used, Available Spare, data written, temperature, unsafe shutdowns, and media errors.

This matters when buying used laptops because the same utility screen can mean different things depending on drive type.

Temperature is a health value and a performance value

NVMe drives can run hot, especially in thin laptops with limited airflow. The health log exposes composite temperature and can record time spent above warning or critical thresholds.

High temperature can reduce performance through thermal throttling and can contribute to long-term stress.

Run a sustained test, not only a quick benchmark

Short benchmarks can sit inside an SSD’s fast cache and hide weak sustained behavior.

For a used laptop, copy a large file set or run a reputable sustained-storage test while monitoring temperature and transfer rate. You are looking for stability, not the highest screenshot number.

Some low-cost or nearly full drives slow sharply once cache is exhausted. That is a performance characteristic, not necessarily a health failure, but it can still affect whether the laptop feels good to use.

Free space changes SSD behavior

SSDs use free flash space for wear leveling, background cleanup, and caching. A nearly full SSD can perform worse than the same SSD with comfortable free space.

Before condemning a used drive for poor performance, check whether it is almost full. But do not erase the machine before collecting health information—you want to inspect the drive’s current state first.

Does formatting or reinstalling Windows reset SSD wear?

No. Reinstalling Windows can remove software clutter and filesystem problems, but it does not restore consumed flash endurance.

Hardware health counters live at the drive/controller level, not inside your Windows installation.

This is similar to TechReviewz’s broader point in Why Phones Slow Down With Age—and What a Factory Reset Cannot Fix: software cleanup can improve one class of problems while leaving hardware aging unchanged.

High SSD health does not make a used laptop a safe purchase by itself

• a worn battery;
• a damaged charging port;
• liquid history;
• thermal throttling;
• a cracked hinge;
• firmware or account locks;
• non-upgradable RAM that is already insufficient.

TechReviewz’s used-laptop inspection checklist covers the rest of the machine.

Replaceability matters as much as remaining life

An SSD showing moderate wear is much less worrying in a laptop where the drive is a standard replaceable M.2 module.

The same wear is more important in a device where storage is soldered or difficult to service. Before buying, identify the exact laptop model and storage layout. “NVMe SSD” does not automatically mean “easy to replace.”

Worked example: 94% health, zero media errors

A three-year-old laptop shows roughly 6% endurance used, healthy spare capacity, no critical warnings, no media errors, moderate power-on hours, and stable sustained performance.

That is generally a reassuring storage profile. The drive has been used, but nothing in the health log points to immediate trouble. You would still inspect the rest of the laptop and confirm the SSD model and replaceability.

Worked example: 99% health but media errors are increasing

Another laptop shows 99% remaining life. At first glance it looks nearly new.

However, the media-error count is non-zero and increases after testing. The system has also logged reliability warnings. The headline health number should not override those stronger warning signs.

Worked example: 70% health on a replaceable SSD

A business laptop has consumed roughly 30% of its estimated endurance but otherwise reports no warnings or media errors. The SSD is a standard M.2 module and replacement cost is modest.

This may still be a perfectly reasonable purchase if the price reflects the wear and the rest of the machine is strong. Used-device value depends on repair cost and replaceability, not only on whether every component is near-new.

A practical SSD checklist for used laptops

1. Identify the exact SSD model and capacity.
2. Check NVMe Percentage Used or the equivalent endurance indicator.
3. Check Critical Warning.
4. Check Available Spare and its threshold.
5. Check media/data-integrity errors.
6. Check temperature and any warning-temperature history.
7. Check power-on hours and unsafe shutdowns.
8. Check host data written.
9. Run a sustained transfer or benchmark.
10. Confirm whether the SSD can be replaced economically.

What should make you walk away?

Strong warning signs include active reliability warnings, read-only state, very low spare capacity, increasing media errors, repeated crashes linked to storage, or a soldered storage design where replacement would be impractical.

A single imperfect number is not always enough to reject a machine. Multiple failure-related signals together are different.

The number to remember is not health percentage

The best used-device decision does not come from one green percentage in a utility.

Read the SSD like a history: how much endurance has been used, how long it has been powered, how much has been written, whether spare capacity is healthy, whether errors exist, how hot it runs, and whether it can be replaced.

Updated October 10, 2026: added a practical application and decision record. This documentation-based guide uses estimates where stated; individual condition, repair costs and regional support vary.